Conducting the Count
Part of Census and Demographics
The step-by-step operational management of a census field campaign, from preparation through data collection to submission.
Why This Matters
The best questionnaire design and the most thorough training are wasted if the field operation is disorganized. Census enumeration is a logistics challenge — coordinating dozens or hundreds of people across a wide area, on a tight schedule, to collect consistent data from thousands of households. Small operational failures multiply: an enumerator who misunderstands one question produces hundreds of incorrect answers; a settlement that receives no visit appears not to exist; forms that are lost before reaching the office destroy weeks of field work.
This article covers the operational discipline needed to execute a census field campaign successfully. It assumes the design is finalized and training is complete; it focuses on the management and quality control systems that turn trained field workers and a good questionnaire into reliable population data.
Pre-Campaign Preparation
Area and workload assignment: Divide the census area into segments small enough for one enumerator to complete in the allotted time. A typical assignment for a household survey might be 30–50 households per enumerator per day, depending on household density, terrain, and interview length. Assign segments on paper (or map) before deploying anyone, with clear geographic boundaries for each assignment.
Household listing: Before the main questionnaire interview, create a complete list of all dwellings in each segment. This is the sampling frame — the list from which households are visited. It prevents “convenient” enumeration (visiting only accessible households and missing remote or hidden ones). The listing can be done during a separate preparatory sweep or during the first day of enumeration.
Materials distribution: Each enumerator should receive:
- Questionnaire forms (more than they expect to need — always run out)
- A map or sketch showing their assignment boundaries
- A completed household listing for their area (if pre-listed)
- Instructions booklet
- Supervisor contact information
- Census ID card or letter of authorization
Logistics: Confirm housing, food, and transport for enumerators working away from their home base. An enumerator who is cold, hungry, or uncertain about getting home cuts interviews short and makes errors. These logistics failures are among the most common causes of poor data quality in field campaigns.
Daily Field Operations
Daily briefing: Before each day’s work, supervisors meet with their teams briefly to review the previous day’s work, address questions, and preview the current day’s assignment. This meeting catches problems early and prevents them from propagating through an entire enumerator-day of work.
Interview sequence: At each household:
- Introduce yourself, your organization, and the purpose of the census.
- Confirm the household identity (address or household listing number).
- Identify the appropriate respondent (usually the household head or another adult).
- Proceed through the questionnaire in order.
- Thank the respondent and note any follow-up items needed.
- Mark the household as completed on your listing.
Non-contact procedure: If no one is home, note the visit time and return at a different time of day. Make at least three attempts at different times before classifying a household as “non-contact.” Record all attempts in the field notes.
Refusals: If a household refuses, note this clearly, do not argue, and inform your supervisor. Some refusals are resolved through supervisor callbacks; others are irreducible. Record refusals honestly rather than inventing data.
Form completion: Complete each form fully, in the field, before leaving the household. Never plan to “fill in the details later” — memory fades, forms get mixed up, and errors are introduced. Write legibly. Use the coding conventions consistently.
Supervision and Field Quality Control
Supervisors are the most important element of field quality control. They should:
Check completed forms daily: Review each enumerator’s forms for completeness (no blank fields without a valid code), consistency (ages consistent with stated relationships — a 20-year-old cannot have a 40-year-old child), and plausibility (a household of 35 people is unusual; verify it).
Re-interview a sample: For each enumerator, select 2–3 completed households at random each day and re-interview them. Compare results with the original form. Consistent discrepancies reveal systematic interview errors or falsification. Small random differences are expected.
Accompany enumerators: Periodically sit in on interviews to observe technique. Provide feedback on areas of improvement. An enumerator who has never been observed may have developed idiosyncratic interview patterns that produce systematically biased data.
Track coverage: Maintain a daily running tally of households listed, households visited, completed interviews, refusals, and non-contacts. This tracking reveals areas falling behind schedule before the problem is irreversible.
Form Collection and Custody
Completed forms are irreplaceable. Protect them from loss, damage, and unauthorized access.
Daily submission: Enumerators submit completed forms to their supervisor each evening (or at a specified interval). Forms should not accumulate in enumerators’ possession for more than 2–3 days.
Supervisors to central office: Supervisors batch forms and submit to the central office with a transmittal record listing how many forms from which segments. The central office signs the transmittal, confirming receipt.
Secure storage: Forms await data entry in a locked, dry, secure location. Access is restricted to processing staff. Any loss, damage, or unauthorized access is documented and reported.
Backup copies: For critical censuses, photograph or copy key forms before shipping. The risk of a messenger losing a batch of forms is real; a backup in the sending office allows reconstruction.
Closing Out the Field Campaign
The campaign is not complete when the last interview is conducted. It is complete when:
- All segment assignments are confirmed completed or documented as incomplete with reasons
- All outstanding non-contacts have been resolved or formally recorded as non-contact
- All supervisors have confirmed their batch totals match their transmittals to the central office
- Central office has confirmed receipt of all expected batches
Final coverage report: The supervisors collectively produce a coverage report: how many households were listed, how many interviewed (by completion status), and the overall response rate. This report is the quality documentation accompanying the dataset — it tells analysts how complete the data is before they work with it.
An undocumented coverage shortfall is as dangerous as the shortfall itself: planners who assume 100% coverage when coverage was actually 80% will make systematically incorrect decisions.